The ability to control bioprocesses at their optimal states is of considerable interest to modern fermentation industries. The aim is to reduce production costs and increase yield while at the same time maintaining the quality of the metabolic products. The objetive of this paper is to apply fuzzy logic to the Penicillin G acylase (PGA) production process. PGA hydrolyzes penicillin G to yield 6-aminopenicilanic acid (6-APA) and phenyl acetic acid. 6-APA is an important raw material, used to produce semi-synthetic ?-lactam antibiotics. Ten experiments were carried out in a bench-scale bioreactor with diferent experimental conditions. Several variables were tested to identify the best fuzzy inputs. The results show that batch time, carbon dioxide concentration and its derivate were the best choice. The algorithm employs these on-line measurements to infer the optimal interval to stop the cultivation. The fuzzy algorithm could accurately infer the time for bioreactor harvesting and the output linguistic variable “stop cultivation” was defined between 0 and 100% certainty. The best experimental result was obtained when the dissolved oxygen concentration was maintained close to 0.5% of saturation.